International Conference on Information and Knowledge Technology

فارسی

Home / چهاردهمین کنفرانس بین المللی فناوری اطلاعات و دانش

Enhancing Software Effort Estimation with an Integrated Approach of Particle Swarm Optimization and Genetic Algorithms in Analogy-based Method

Authors :

Ehsan Nasr¹ Keyvan Mohebbi²

1- دانشگاه آزاد اسلامی واحد اصفهان (خوراسگان) 2- دانشگاه آزاد اسلامی واحد اصفهان (خوراسگان)

Keywords :

Software Effort Estimation،Analogy-based Estimation،Non-algorithmic Model،Genetic Algorithm،Particle Swarm Optimization

Abstract :

Analogy-based estimation is a widely used approach for software effort estimation that involves comparing a new project to similar completed projects. However, this method may not work effectively when there are variations in the importance of project features or dependencies between them. To address this issue, features can be assigned weights using optimization techniques, such as meta-heuristic algorithms. Nevertheless, these algorithms may become stuck in a local optimum, leading to suboptimal results. This research aims to obtain the global optimal weights for project features by combining particle swarm and genetics algorithms. This hybrid algorithm utilizes the motion of particles in the state space and the composition and mutation of particles to generate more potential solutions, thus increasing the chance of finding the global optimal and overcoming the local optimal problem. With this algorithm, the weights for the project features are calculated and utilized to estimate the software development effort using the analogy-based method. The proposed approach was tested and assessed using two datasets, namely Maxwell and Desharnais. The experimental results indicated an enhancement in the evaluation criteria, including MMRE, MdMRE, and PRED. The TotalCost, which incorporates all these criteria, was notably improved by the proposed approach in both the Maxwell and Desharnais datasets, achieving a 75.894% and 107.246% increase, respectively, compared to prior research.